This project is concerned with statistical methodology issues involved in the design and analysis of animal experiments with particular emphasis on TRTP's two-year carcinogenesis studies. Research activities fall into three general categories: experimental design, statistical methodology development and data base analyses. Specific problems include: (1) modification of the experimental design employed in TRTP carcinogenicity studies, (2) use of logistic regression in the analysis of tumor incidence, (3) development of statistical procedures for incorporating historical control data in a formal testing framework, and (4) estimation of the false positive rate in two-year carcinogenicity experiments.